Commit Graph

71569 Commits

Author SHA1 Message Date
Eric Christopher de9d58569f Add more comments... err debug statements to the fast allocator.
llvm-svn: 129400
2011-04-12 22:17:44 +00:00
Oscar Fuentes 80cd17c79f Fix compiler command line used by lit.py when working with NMake
generators. It may improve robustness when testing from VS too.

Based on a patch by David Neto!

llvm-svn: 129398
2011-04-12 22:10:38 +00:00
Johnny Chen 960eef3db3 The Thumb2 RFE instructions need to have their second halfword fully specified.
In addition, the base register is not rGPR, but GPR with th exception that:

    if n == 15 then UNPREDICTABLE

rdar://problem/9273836

llvm-svn: 129391
2011-04-12 21:41:51 +00:00
Jakob Stoklund Olesen c49df2c05a SparseBitVector is SLOW.
Use a Bitvector instead, we didn't need the smaller memory footprint anyway.
This makes the greedy register allocator 10% faster.

llvm-svn: 129390
2011-04-12 21:30:53 +00:00
Jim Grosbach 733d305fee MCJIT lazy relocation resolution and symbol address re-assignment.
Add handling for tracking the relocations on symbols and resolving them.
Keep track of the relocations even after they are resolved so that if
the RuntimeDyld client moves the object, it can update the address and any
relocations to that object will be updated.

For our trival object file load/run test harness (llvm-rtdyld), this enables
relocations between functions located in the same object module. It should
be trivially extendable to load multiple objects with mutual references.

As a simple example, the following now works (running on x86_64 Darwin 10.6):


$ cat t.c
int bar() {
  return 65;
}

int main() {
  return bar();
}
$ clang t.c -fno-asynchronous-unwind-tables -o t.o -c
$ otool -vt t.o
t.o:
(__TEXT,__text) section
_bar:
0000000000000000  pushq %rbp
0000000000000001  movq  %rsp,%rbp
0000000000000004  movl  $0x00000041,%eax
0000000000000009  popq  %rbp
000000000000000a  ret
000000000000000b  nopl  0x00(%rax,%rax)
_main:
0000000000000010  pushq %rbp
0000000000000011  movq  %rsp,%rbp
0000000000000014  subq  $0x10,%rsp
0000000000000018  movl  $0x00000000,0xfc(%rbp)
000000000000001f  callq 0x00000024
0000000000000024  addq  $0x10,%rsp
0000000000000028  popq  %rbp
0000000000000029  ret
$ llvm-rtdyld t.o -debug-only=dyld ; echo $?
Function sym: '_bar' @ 0
Function sym: '_main' @ 16
Extracting function: _bar from [0, 15]
    allocated to 0x100153000
Extracting function: _main from [16, 41]
    allocated to 0x100154000
Relocation at '_main' + 16 from '_bar(Word1: 0x2d000000)
Resolving relocation at '_main' + 16 (0x100154010) from '_bar (0x100153000)(pcrel, type: 2, Size: 4).
loaded '_main' at: 0x100154000
65
$

llvm-svn: 129388
2011-04-12 21:20:41 +00:00
Johnny Chen 01637b9acb Add bad register checks for Thumb2 Ld/St instructions.
rdar://problem/9269047

llvm-svn: 129387
2011-04-12 21:17:51 +00:00
Andrew Trick 1b60ad6644 Revert 129383. It causes some targets to hit a scheduler assert.
llvm-svn: 129385
2011-04-12 20:14:07 +00:00
Nick Lewycky 2b473f8c85 Print our uint64_t with the more portable (C99 and C++0x) %PRIu64 format
specifier.

llvm-svn: 129384
2011-04-12 20:06:50 +00:00
Andrew Trick c5dd24a542 PreRA scheduler heuristic fixes: VRegCycle, TokenFactor latency.
UnitsSharePred was a source of randomness in the scheduler: node
priority depended on the queue data structure. I rewrote the recent
VRegCycle heuristics to completely replace the old heuristic without
any randomness. To make these heuristic adjustments to node latency work,
I also needed to do something a little more reasonable with TokenFactor. I
gave it zero latency to its consumers and always schedule it as low as
possible.

llvm-svn: 129383
2011-04-12 19:54:36 +00:00
Oscar Fuentes 23fe31cc8d Document how to build a LLVM pass with CMake out of source.
Patch by arrowdodger!

llvm-svn: 129381
2011-04-12 19:40:35 +00:00
Nick Lewycky 3c974a6bde The counters are unsigned.
llvm-svn: 129380
2011-04-12 19:35:45 +00:00
Jakob Stoklund Olesen c70b697a40 Create new intervals for isolated blocks during region splitting.
This merges the behavior of splitSingleBlocks into splitAroundRegion, so the
RS_Region and RS_Block register stages can be coalesced. That means the leftover
intervals after region splitting go directly to spilling instead of a second
pass of per-block splitting.

llvm-svn: 129379
2011-04-12 19:32:53 +00:00
Rafael Espindola fd794affe5 Remove LastOffset from the asm parser.
llvm-svn: 129378
2011-04-12 18:53:30 +00:00
Johnny Chen ab86a519f8 The Thumb2 Ld, St, and Preload instructions with the i12 forms should have its Inst{23}
be specified as '1' (add = TRUE).

Also add a utility function for Thumb2.

llvm-svn: 129377
2011-04-12 18:48:00 +00:00
Jakob Stoklund Olesen 0840f50b76 Add SplitKit API to query and select the current interval being worked on.
This makes it possible to target multiple registers in one pass.

llvm-svn: 129374
2011-04-12 18:11:31 +00:00
Jakob Stoklund Olesen 68e84581c5 Fix a bug in RegAllocBase::addMBBLiveIns() where a basic block could accidentally be skipped.
llvm-svn: 129373
2011-04-12 18:11:28 +00:00
Devang Patel 4547a9e658 Remove dead typedef.
llvm-svn: 129368
2011-04-12 17:43:12 +00:00
Devang Patel 5eb4319dba Refactor CompileUnit into a separate header.
llvm-svn: 129367
2011-04-12 17:40:32 +00:00
Johnny Chen d0e2be39ea Print out a debug message when the reglist fails the sanity check for Thumb Ld/St Multiple.
llvm-svn: 129365
2011-04-12 17:09:04 +00:00
Rafael Espindola 1ec0f46169 Fix the case of a .cfi_rel_offset before any .cfi_def_cfa_offset.
llvm-svn: 129362
2011-04-12 16:12:03 +00:00
Rafael Espindola 2e1c9d2188 Implement .cfi_same_value.
llvm-svn: 129361
2011-04-12 15:31:05 +00:00
Garrison Venn a0f6ecb01f Added new FIXME note
llvm-svn: 129360
2011-04-12 12:30:10 +00:00
Cameron Zwarich fbcd69b96a Split a store of a VMOVDRR into two integer stores to avoid mixing NEON and ARM
stores of arguments in the same cache line. This fixes the second half of
<rdar://problem/8674845>.

llvm-svn: 129345
2011-04-12 02:24:17 +00:00
NAKAMURA Takumi 3f28443a07 lib/Transforms/Instrumentation/CMakeLists.txt: Add LineProfiling.cpp to fix up r129340.
llvm-svn: 129343
2011-04-12 01:54:40 +00:00
Nick Lewycky 9d60e373cf Add support for line profiling. Very work-in-progress.
Use debug info in the IR to find the directory/file:line:col. Each time that location changes, bump a counter.

Unlike the existing profiling system, we don't try to look at argv[], and thusly don't require main() to be present in the IR. This matches GCC's technique where you specify the profiling flag when producing each .o file.

The runtime library is minimal, currently just calling printf at program shutdown time. The API is designed to make it possible to emit GCOV data later on.

llvm-svn: 129340
2011-04-12 01:06:09 +00:00
Nick Lewycky fbc5a4004c Consider ConstantAggregateZero as well as ConstantArray/Struct.
llvm-svn: 129338
2011-04-12 01:02:45 +00:00
Eric Christopher c37833625a Fix typo.
llvm-svn: 129334
2011-04-12 00:48:08 +00:00
Nick Lewycky 11168326f8 Make IRBuilder support StringRef for building strings.
Also document that the global variables produced are mergable.

llvm-svn: 129330
2011-04-12 00:29:07 +00:00
Jim Grosbach 3ed03f18d1 Tidy up a bit now that we're using the MemoryManager interface.
llvm-svn: 129328
2011-04-12 00:23:32 +00:00
Johnny Chen f130b7f0f5 Add one test case (svc).
llvm-svn: 129327
2011-04-12 00:21:48 +00:00
Eric Christopher ffc0e1f6e6 Match case for invalid constant error messages and add a new
test for invalid hexadecimals.

llvm-svn: 129326
2011-04-12 00:18:03 +00:00
Johnny Chen 672ef14a62 A8.6.16 B
Encoding T1 (tBcc)
if cond == '1110' then UNDEFINED;

rdar://problem/9268681

llvm-svn: 129325
2011-04-12 00:14:49 +00:00
Dan Gohman 1c6c34834b Fix reassociate to use a worklist instead of recursing when new
reassociation opportunities are exposed. This fixes a bug where
the nested reassociation expects to be the IR to be consistent,
but it isn't, because the outer reassociation has disconnected
some of the operands.  rdar://9167457

llvm-svn: 129324
2011-04-12 00:11:56 +00:00
Eric Christopher 0a68959604 Test for invalid constant expr addition - bad octal constant.
llvm-svn: 129323
2011-04-12 00:03:38 +00:00
Eric Christopher 104af0619e To avoid printing out multiple error messages for cases like:
.long 80+08

go ahead and assume that if we've got an Error token that we handled it
already. Otherwise if it's a token we can't handle then go ahead and
return the default error.

llvm-svn: 129322
2011-04-12 00:03:13 +00:00
Jakob Stoklund Olesen 507992e909 Reuse live interval union between functions. This saves a bit of compile time
when compiling many small functions.

llvm-svn: 129321
2011-04-11 23:57:14 +00:00
Johnny Chen dc8bf9ec08 Thumb disassembler was erroneously rejecting "blx sp" instruction.
rdar://problem/9267838

llvm-svn: 129320
2011-04-11 23:33:30 +00:00
Chris Lattner 7d4cdae564 comment cleanup, use moveBefore instead of removeFromParent+insertBefore.
llvm-svn: 129319
2011-04-11 23:24:57 +00:00
NAKAMURA Takumi 408c228367 TableGen: Keep the order of DECL_CONTEXT() for DeclNodes.td. RecordVector may be used instead of RecordSet.
The result of DeclNodes.inc was unstable on msys, Windows 7 x64.

llvm-svn: 129317
2011-04-11 23:20:28 +00:00
Chris Lattner e81d045d94 remove the StructRetPromotion pass. It is unused, not maintained and
has some bugs.  If this is interesting functionality, it should be 
reimplemented in the argpromotion pass.

llvm-svn: 129314
2011-04-11 23:09:44 +00:00
Wesley Peck f30a0e2d80 Fix an error in the MBlaze delay slot filler.
llvm-svn: 129313
2011-04-11 22:45:02 +00:00
Daniel Dunbar 0625b1f860 build: Add support for a SHOW_DIAGNOSTICS build variable.
If enabled, this will attempt to use the CC_LOG_DIAGNOSTICS feature I dropped
into Clang to print a log of all the diagnostics generated during an individual
build (from the top-level). Not sure if this will actually be useful, but for
now it is handy for testing the option.

llvm-svn: 129312
2011-04-11 22:37:39 +00:00
Wesley Peck 1914c39bd4 Add scheduling information for the MBlaze backend.
llvm-svn: 129311
2011-04-11 22:31:52 +00:00
Eric Christopher 64749f2a89 Lex, and then fail on invalid constants.
Testcase forthcoming.

rdar://8490596

llvm-svn: 129309
2011-04-11 22:24:56 +00:00
Nick Lewycky 0f85789800 Just because a GlobalVariable's initializer is [N x { i32, void ()* }] doesn't
mean that it has to be ConstantArray of ConstantStruct. We might have
ConstantAggregateZero, at either level, so don't crash on that.

Also, semi-deprecate the sentinal value. The linker isn't aware of sentinals so
we end up with the two lists appended, each with their "sentinals" on them.
Different parts of LLVM treated sentinals differently, so make them all just
ignore the single entry and continue on with the rest of the list.

llvm-svn: 129307
2011-04-11 22:11:20 +00:00
Rafael Espindola 82065cb6cf Implement cfi_rel_offset
llvm-svn: 129306
2011-04-11 21:49:50 +00:00
Jakob Stoklund Olesen 0f175ebc32 Speed up eviction by stopping collectInterferingVRegs as soon as the spill
weight limit has been exceeded.

llvm-svn: 129305
2011-04-11 21:47:01 +00:00
Rafael Espindola 4f46abc15e Add test for previous commit.
llvm-svn: 129304
2011-04-11 21:41:34 +00:00
Wesley Peck e3685217d0 Don't crash on invalid instructions when disassembling MBlaze code.
This fixes http://llvm.org/bugs/show_bug.cgi?id=9653

llvm-svn: 129303
2011-04-11 21:35:21 +00:00
Bill Wendling 1e1f1c9ce1 The default of the dispatch switch statement was to branch to a BB that executed
the 'unwind' instruction. However, later on that instruction was converted into
a jump to the basic block it was located in, causing an infinite loop when we
get there.

It turns out, we get there if the _Unwind_Resume_or_Rethrow call returns (which
it's not supposed to do). It returns if it cannot find a place to unwind
to. Thus we would get what appears to be a "hang" when in reality it's just that
the EH couldn't be propagated further along.

Instead of infinitely looping (or calling `unwind', which none of our back-ends
support (it's lowered into nothing...)), call the @llvm.trap() intrinsic
instead. This may not conform to specific rules of a particular language, but
it's rather better than infinitely looping.

<rdar://problem/9175843&9233582>

llvm-svn: 129302
2011-04-11 21:32:34 +00:00
Johnny Chen f79d5365de Fix the bug where the immediate shift amount for Thumb logical shift instructions are incorrectly disassembled.
rdar://problem/9266265

llvm-svn: 129298
2011-04-11 21:14:35 +00:00
Evan Cheng ef42bea704 Look pass copies when determining whether hoisting would end up inserting more copies. rdar://9266679
llvm-svn: 129297
2011-04-11 21:09:18 +00:00
Rafael Espindola ffd2e5163b implement .cfi_adjust_cfa_offset.
llvm-svn: 129296
2011-04-11 20:29:16 +00:00
Eli Friedman 2c81975bae Get rid of useless comment; if a file uses functions from a given header,
it is obvious that it should be included.

llvm-svn: 129295
2011-04-11 20:16:05 +00:00
Owen Anderson 5140802cd9 Fix another using-CPSR-twice bug in my ADCS/SBCS cleanups, and make proper use of the Commutable bit.
llvm-svn: 129294
2011-04-11 20:12:19 +00:00
Jakob Stoklund Olesen b16e948029 Skip a binary search when possible.
llvm-svn: 129293
2011-04-11 20:01:44 +00:00
Jakob Stoklund Olesen 7d05bce70c Use a faster algorithm for computing MBB live-in registers after register allocation.
LiveIntervals::findLiveInMBBs has to do a full binary search for each segment.

llvm-svn: 129292
2011-04-11 20:01:41 +00:00
Garrison Venn 56c5ca2ee1 Because some systems have reported that this example would not build the
header file cstdio was added as an include.

llvm-svn: 129291
2011-04-11 19:52:49 +00:00
Johnny Chen 74adbddade Trivial comment fix.
llvm-svn: 129288
2011-04-11 18:51:50 +00:00
Evan Cheng fe917efc8b Fix a couple of places where changes are made but not tracked.
llvm-svn: 129287
2011-04-11 18:47:20 +00:00
Johnny Chen 66fab75920 Check invalid register encodings for LdFrm/StFrm ARM instructions and flag them as
invalid instructions.

llvm-svn: 129286
2011-04-11 18:34:12 +00:00
Kevin Enderby 9377a52c12 Adding support for printing operands symbolically to llvm's public 'C'
disassembler API.  Hooked this up to the ARM target so such tools as Darwin's
otool(1) can now print things like branch targets for example this:
  blx _puts
instead of this:
  blx #-36
And even print the expression encoded in the Mach-O relocation entried for
things like this:
  movt r0, :upper16:((_foo-_bar)+1234)

llvm-svn: 129284
2011-04-11 18:08:50 +00:00
Jakob Stoklund Olesen f8beafe207 Don't add live ranges for sub-registers when clobbering a physical register.
Both coalescing and register allocation already check aliases for interference,
so these extra segments are only slowing us down.

This speeds up both linear scan and the greedy register allocator.

llvm-svn: 129283
2011-04-11 18:08:10 +00:00
Jakob Stoklund Olesen 4fbbe3689d Speed up LiveIntervalUnion::unify by handling end insertion specially.
This particularly helps with the initial transfer of fixed intervals.

llvm-svn: 129277
2011-04-11 15:00:44 +00:00
Jakob Stoklund Olesen bfabc494f5 Time the initial seeding of live registers
llvm-svn: 129276
2011-04-11 15:00:42 +00:00
Jakob Stoklund Olesen 96d04c8e00 Don't shrink live ranges after dead code elimination unless it is going to help.
In particular, don't repeatedly recompute the PIC base live range after rematerialization.

llvm-svn: 129275
2011-04-11 15:00:39 +00:00
Oscar Fuentes c6f2d0a4cf CMake: remove some unnecesary code and ToDos.
Patch by arrowdodger!

llvm-svn: 129274
2011-04-11 14:52:39 +00:00
Jay Foad 0159a1ee11 Fix or remove code which seemed to think that the operand of a Constant
was always a User.

llvm-svn: 129272
2011-04-11 09:48:55 +00:00
Jay Foad 7c14a558fe Don't include Operator.h from InstrTypes.h.
llvm-svn: 129271
2011-04-11 09:35:34 +00:00
Jay Foad 29426e87c1 Phi nodes always use an even number of operands, so don't ever allocate
an odd number.

llvm-svn: 129270
2011-04-11 09:25:51 +00:00
NAKAMURA Takumi 62920834fa docs/CommandGuide/index.html: Fix CSS url to relative.
llvm-svn: 129268
2011-04-11 03:27:38 +00:00
NAKAMURA Takumi 2a7024fa6c docs/CommandLine.html: &quot;ize.
llvm-svn: 129267
2011-04-11 03:27:31 +00:00
Bill Wendling 35a9c3cd72 Revert r129235 pending a vetting of the EH rewrite.
--- Reverse-merging r129235 into '.':
D    test/Feature/bb_attrs.ll
U    include/llvm/BasicBlock.h
U    include/llvm/Bitcode/LLVMBitCodes.h
U    lib/VMCore/AsmWriter.cpp
U    lib/VMCore/BasicBlock.cpp
U    lib/AsmParser/LLParser.cpp
U    lib/AsmParser/LLLexer.cpp
U    lib/AsmParser/LLToken.h
U    lib/Bitcode/Reader/BitcodeReader.cpp
U    lib/Bitcode/Writer/BitcodeWriter.cpp

llvm-svn: 129259
2011-04-10 23:18:04 +00:00
Nicolas Geoffray 9137ee85de Bugfix in the Cpp backend after API change on PHINode::Create.
llvm-svn: 129248
2011-04-10 17:39:40 +00:00
Oscar Fuentes dbe99ba37d CMake: support for using LLVM from client projects with find_package.
Patch by arrowdodger!

llvm-svn: 129247
2011-04-10 16:17:49 +00:00
Garrison Venn 88bd9d6bb8 Fixed more best practices, and explicit/implicit style guide issues.
llvm-svn: 129245
2011-04-10 14:06:52 +00:00
Oscar Fuentes fe40844fb6 Reword instructions for running tests on Visual Studio.
Suggested by Maël Hörz.

llvm-svn: 129237
2011-04-10 01:41:55 +00:00
Bill Wendling 3d5450d809 Beginning of the Great Exception Handling Rewrite.
* Add a "landing pad" attribute to the BasicBlock.
* Modify the bitcode reader and writer to handle said attribute.

Later: The verifier will ensure that the landing pad attribute is used in the
appropriate manner. I.e., not applied to the entry block, and applied only to
basic blocks that are branched to via a `dispatch' instruction.

(This is a work-in-progress.)

llvm-svn: 129235
2011-04-10 00:04:27 +00:00
Chris Lattner b799b2d4be drive by bug fix: DenseMapInfo::isEqual should be determined according to
isEqual of its members, not operator==.

llvm-svn: 129233
2011-04-09 21:20:23 +00:00
Chris Lattner 077f0179bb LLVMCC_EMITIR_FLAG is already either -emit-llvm or -fplugin-arg-dragonegg-emit-ir, so don't add an extra leading -. Clang doesn't accept --emit-llvm.
llvm-svn: 129232
2011-04-09 19:56:28 +00:00
Chris Lattner 96d2507e79 make stuff work when LLVMCC_OPTION is something like installed-clang, not just clang.
llvm-svn: 129229
2011-04-09 19:53:16 +00:00
Chris Lattner fc4fe00a65 fix rdar://8735979 - "int 3" doesn't match to "int3". Unfortunately,
InstAlias doesn't allow matching immediate operands, so we have to write
C++ code to do this.

llvm-svn: 129223
2011-04-09 19:41:05 +00:00
Chris Lattner 214f114aa7 look for the verboten argument slot access in any order, thanks to Frits
for pointing this out

llvm-svn: 129217
2011-04-09 17:00:34 +00:00
Benjamin Kramer 26e7768f38 Fix potential buffer overflow on win32.
llvm-svn: 129214
2011-04-09 14:06:12 +00:00
Benjamin Kramer 2b6c96b43d Don't store Twine temporaries, it's not safe.
And don't append the name over and over again in the loop.

llvm-svn: 129210
2011-04-09 11:26:27 +00:00
Benjamin Kramer 2792f2b86b Make error message more useful.
llvm-svn: 129209
2011-04-09 10:10:35 +00:00
NAKAMURA Takumi 6bd36d56e4 docs/*.html: Make W3C HTML 4.01 Strict more compliant.
FIXME: The logo handling in ReleaseNotes.html
llvm-svn: 129208
2011-04-09 09:51:57 +00:00
Eli Friedman 9cca0715aa Add back a couple checks removed by r129128; the fact that an intitializer
is an array of structures doesn't imply it's a ConstantArray of
ConstantStruct.

llvm-svn: 129207
2011-04-09 09:11:09 +00:00
Chris Lattner 88974f4625 fix PR9523, a crash in looprotate on a non-canonical loop made out of indirectbr.
llvm-svn: 129203
2011-04-09 07:25:58 +00:00
Chris Lattner af1bccec68 Fix a bug where RecursivelyDeleteTriviallyDeadInstructions could
delete the instruction pointed to by CGP's current instruction
iterator, leading to a crash on the testcase.  This fixes PR9578.

llvm-svn: 129200
2011-04-09 07:05:44 +00:00
Chris Lattner cfe5aa65d2 Avoid excess precision issues that lead to generating host-compiler-specific code.
Switch lowering probably shouldn't be using FP for this.  This resolves PR9581.

llvm-svn: 129199
2011-04-09 06:57:13 +00:00
Eli Friedman 17822fcde9 PR9604; try to deal with RAUW updates correctly in the AST. I'm not convinced
it's completely safe to cache the AST across LICM runs even with this fix,
but this fix can't hurt.

llvm-svn: 129198
2011-04-09 06:55:46 +00:00
Eli Friedman 4db39cefdb Test for r129190.
llvm-svn: 129197
2011-04-09 06:39:43 +00:00
Chris Lattner e53c95f180 fix PR9629 - We were lowering regexes like a{{b|c}}d into ab|cd, which
is substantially different than a(b|c)d.  Form the latter regex instead.
This found a few problems in the testsuite, which serves as its test.

llvm-svn: 129196
2011-04-09 06:37:03 +00:00
Chris Lattner 418b1037b0 fix two completely broken tests, which were matching due to PR9629.
llvm-svn: 129195
2011-04-09 06:34:38 +00:00
Chris Lattner ea6afab4b0 remove a bunch of CHECK lines that aren't checking what
they thought they were, because alternation was expanding
wrong in {{}}'s.

llvm-svn: 129194
2011-04-09 06:31:06 +00:00
Chris Lattner 0687ec7051 regexes are allowed to match empty things, e.g. {{.*}} in filecheck.
llvm-svn: 129193
2011-04-09 06:29:24 +00:00
Chris Lattner 53e0679d6f various cleanups, no functionality change.
llvm-svn: 129192
2011-04-09 06:18:02 +00:00
Chris Lattner 8872ec3ff0 revert accidental commit.
llvm-svn: 129191
2011-04-09 06:01:28 +00:00
Chris Lattner 2bd7015de5 fix a potentially serious bug in AliasSet::removeCallSite
where we shrunk the list without updating the end iterator.
By inspection, from PR9639.  

llvm-svn: 129190
2011-04-09 05:51:34 +00:00
Jakob Stoklund Olesen ed47ed4e80 Build the Hopfield network incrementally when splitting global live ranges.
It is common for large live ranges to have few basic blocks with register uses
and many live-through blocks without any uses. This approach grows the Hopfield
network incrementally around the use blocks, completely avoiding checking
interference for some through blocks.

llvm-svn: 129188
2011-04-09 02:59:09 +00:00
Jakob Stoklund Olesen 4ad6c160a5 Precompute interference for neighbor blocks as long as there is no interference.
This doesn't require seeking in the live interval union, so it is very cheap.

llvm-svn: 129187
2011-04-09 02:59:05 +00:00
Chris Lattner 0359101df2 fix doc comment bug, noticed by Jochen
llvm-svn: 129186
2011-04-09 02:33:29 +00:00
Chris Lattner 41c80e89f3 have dag combine zap "store undef", which can be formed during call lowering
with undef arguments.

llvm-svn: 129185
2011-04-09 02:32:02 +00:00
Chris Lattner 1c42a4d159 don't test for codegen of 'store undef'
llvm-svn: 129184
2011-04-09 02:31:26 +00:00
NAKAMURA Takumi e8a8fd658f docs/tutorial/OCamlLangImpl7.html: Fix a href to missing Chapter 8.
llvm-svn: 129183
2011-04-09 02:14:41 +00:00
NAKAMURA Takumi b10df26061 docs: Fix some in-package URLs.
(eg. trim http://llvm.org/docs/foo -> foo)

llvm-svn: 129182
2011-04-09 02:13:48 +00:00
NAKAMURA Takumi ca46f5a3b5 docs: Canonicalize URLs.
llvm-svn: 129181
2011-04-09 02:13:37 +00:00
Sean Callanan 1426351c83 Redirected errors from the AsmParser to the proper
error stream, in cases where the AsmParser is
being invoked by EDDisassembler.  Before, they
were being sent to errs() because no error handler
was installed in the SourceMgr.

llvm-svn: 129177
2011-04-09 00:37:25 +00:00
Sean Callanan 7ccf375622 Moved an access to an object past a NULL check,
making the MC disassembler tester more robust.

llvm-svn: 129175
2011-04-09 00:21:04 +00:00
Devang Patel 12599c2187 Add radar number for future reference.
llvm-svn: 129172
2011-04-08 23:52:04 +00:00
Devang Patel 778947c203 Simplify array bound checks and clarify comments. One element array can have same non-zero number as lower bound as well as upper bound.
llvm-svn: 129170
2011-04-08 23:39:38 +00:00
Nick Lewycky bd10af96bd Add a function for profiling to run at shutdown. Unlike the existing API, this
can be used even when main() isn't present in the Module, but it means that you
don't get to read argv[].

llvm-svn: 129163
2011-04-08 22:19:52 +00:00
Matt Beaumont-Gay 4e1796e8d1 Fix an apparent typo that made GCC complain
llvm-svn: 129160
2011-04-08 21:59:49 +00:00
Devang Patel e39647951b Do not emit DW_AT_upper_bound and DW_AT_lower_bound for unbouded array.
If lower bound is more then upper bound then consider it is an unbounded array.
An array is unbounded if non-zero lower bound is same as upper bound.
If lower bound and upper bound are zero than array has one element.

llvm-svn: 129156
2011-04-08 21:55:10 +00:00
Bill Wendling c5d44ae3f4 Missing closed tag.
llvm-svn: 129155
2011-04-08 21:43:08 +00:00
Evan Cheng 74d92c1924 Change -arm-trap-func= into a non-arm specific option. Now Intrinsic::trap is lowered into a call to the specified trap function at sdisel time.
llvm-svn: 129152
2011-04-08 21:37:21 +00:00
Jim Grosbach ecd7043883 Workaround g++ 4.2.1 warning diagnostic false positive.
llvm-svn: 129149
2011-04-08 21:11:20 +00:00
Johnny Chen f2faf4e53a Check opcoe (dmb, dsb) instead of bitfields matching.
llvm-svn: 129148
2011-04-08 20:03:46 +00:00
Johnny Chen a9570f77d5 Hanlde the checking of bad regs for SMMLAR properly, instead of asserting.
PR9650
rdar://problem/9257565

llvm-svn: 129147
2011-04-08 19:41:22 +00:00
Johnny Chen 875e0e4626 Sanity check the option operand for DMB/DSB.
PR9648
rdar://problem/9257634

llvm-svn: 129146
2011-04-08 19:18:07 +00:00
Jim Grosbach a5dcd98a47 Mark hasExtraDefRegAllocReq=1 on LDRD.
The previous cleanup of LDRD got overzealous and removed it, causing post-RA
scheduling to get overzealous in breaking antidependencies and invalidate these instructions. Hilarity and invalid assembly ensued.

rdar://9244161

llvm-svn: 129144
2011-04-08 18:47:05 +00:00
Chris Lattner a868bbbb66 reindent this whole file and do a variety of stylistic cleanups.
This code is still a long way from following best practices.

llvm-svn: 129140
2011-04-08 18:02:51 +00:00
Chris Lattner ca320f54da fix this to build with the recent StructType changes.
llvm-svn: 129139
2011-04-08 17:56:47 +00:00
Jim Grosbach 18b81c52bb Refactor MCJIT 32-bit section loading.
Teach 32-bit section loading to use the Memory Manager interface, just like
the 64-bit loading does. Tidy up a few other things here and there.

llvm-svn: 129138
2011-04-08 17:31:24 +00:00
Johnny Chen 74c74ba81c MOVi16 and MOVTi16 does not allow pc as the dest register, while MOVi allows it.
Add tests for that.

llvm-svn: 129137
2011-04-08 17:29:58 +00:00
Nick Lewycky 466d0c1f93 llvm.global_[cd]tor is defined to be either external, or appending with an array
of { i32, void ()* }. Teach the verifier to verify that, deleting copies of
checks strewn about.

llvm-svn: 129128
2011-04-08 07:30:21 +00:00
Bill Wendling 33984c6973 Only emit the AvailableFeatures variable if it's used.
llvm-svn: 129124
2011-04-08 04:08:57 +00:00
Bruno Cardoso Lopes a21d470870 Also add the late notes for Mips in the current release note
llvm-svn: 129122
2011-04-08 03:06:22 +00:00
Johnny Chen 7e51b4640f Add sanity checking for bad register specifier(s) for the DPFrm instructions.
Add more test cases to exercise the logical branches related to the above change.

llvm-svn: 129117
2011-04-08 00:29:09 +00:00
Rafael Espindola 9a1a34b105 Update tests
llvm-svn: 129116
2011-04-07 23:51:25 +00:00
Rafael Espindola b80875edee Micro optimization and improved similarity with gas' output:
When two section names share a suffix, reuse the entry in shstrtab.

llvm-svn: 129115
2011-04-07 23:21:52 +00:00
Devang Patel bc3d8b212f Do not let debug info interfer with branch folding.
llvm-svn: 129114
2011-04-07 23:11:25 +00:00
Johnny Chen 165a07adf9 Add a VEXT test.
llvm-svn: 129111
2011-04-07 22:04:01 +00:00
Bill Wendling bc3f79044a Replace the old algorithm that emitted the "print the alias for an instruction"
with the newer, cleaner model. It uses the IAPrinter class to hold the
information that is needed to match an instruction with its alias. This also
takes into account the available features of the platform.

There is one bit of ugliness. The way the logic determines if a pattern is
unique is O(N**2), which is gross. But in reality, the number of items it's
checking against isn't large. So while it's N**2, it shouldn't be a massive time
sink.

llvm-svn: 129110
2011-04-07 21:20:06 +00:00
Rafael Espindola ece7c9c282 If present, use gold's support for getting a file view. This prevents having
to map the file both in the linker and in the plugin.

llvm-svn: 129109
2011-04-07 21:11:00 +00:00
Evan Cheng 9a3f2772f0 Add option to emit @llvm.trap as a function call instead of a trap instruction. rdar://9249183.
llvm-svn: 129107
2011-04-07 20:31:12 +00:00
Rafael Espindola 88bc341061 Add support for .skip.
Patch by Roman Divacky.
Fixes PR9361.

llvm-svn: 129106
2011-04-07 20:26:23 +00:00
Akira Hatanaka 052163e6d3 Fix indentation.
llvm-svn: 129105
2011-04-07 20:25:10 +00:00
Akira Hatanaka 94ee37e487 Update ATUsed every time after expandRegLargeImmPair is called.
llvm-svn: 129104
2011-04-07 20:23:26 +00:00
Mon P Wang 27f3330132 Fixed encoding for VEXTqf
llvm-svn: 129101
2011-04-07 19:56:12 +00:00
Andrew Trick 2ad0b37318 Added a check in the preRA scheduler for potential interference on a
induction variable. The preRA scheduler is unaware of induction vars,
so we look for potential "virtual register cycles" instead.

Fixes <rdar://problem/8946719> Bad scheduling prevents coalescing

llvm-svn: 129100
2011-04-07 19:54:57 +00:00
Akira Hatanaka d6f1c58914 Fix handling of functions with internal linkage.
llvm-svn: 129099
2011-04-07 19:51:44 +00:00
Johnny Chen 04efb8f6ce Add sanity checking for invalid register encodings for signed/unsigned extend instructions.
Add some test cases.

llvm-svn: 129098
2011-04-07 19:28:58 +00:00
Johnny Chen 07606661f9 Add sanity checking for invalid register encodings for saturating instructions.
llvm-svn: 129096
2011-04-07 19:02:08 +00:00
Jakob Stoklund Olesen 64beb47783 Recompute hasPHIKill flags when shrinking live intervals.
PHI values may be deleted, causing the flags to be wrong. This fixes PR9616.

llvm-svn: 129092
2011-04-07 18:43:14 +00:00
Johnny Chen 194a2267ad Add some more comments about checkings of invalid register numbers.
And two test cases.

llvm-svn: 129090
2011-04-07 18:33:19 +00:00
Rafael Espindola e4e4e37580 Expose more passes to the C API.
llvm-svn: 129087
2011-04-07 18:20:46 +00:00
Jakob Stoklund Olesen 994c16833c Avoid moving iterators when the previous block was just visited.
llvm-svn: 129081
2011-04-07 17:27:50 +00:00
Jakob Stoklund Olesen 1c0db0fd21 Prefer multiplications to divisions.
llvm-svn: 129080
2011-04-07 17:27:48 +00:00
Jakob Stoklund Olesen 6d2bbc1c20 Extract SpillPlacement::addLinks for handling the special transparent blocks.
llvm-svn: 129079
2011-04-07 17:27:46 +00:00
Devang Patel 197c35298a While hoisting common code from if/else, hoist debug info intrinsics if they match.
llvm-svn: 129078
2011-04-07 17:27:36 +00:00
Tanya Lattner 266792a55a Prevent ARM DAG Combiner from doing an AND or OR combine on an illegal vector type (vectors of size 3). Also included test cases.
llvm-svn: 129074
2011-04-07 15:24:20 +00:00
Jay Foad 65bef04ea7 Fix a bit of nonsense.
llvm-svn: 129073
2011-04-07 12:41:09 +00:00
Chris Lattner 5ddaab1789 add faust too
llvm-svn: 129053
2011-04-07 03:09:21 +00:00
Chris Lattner caefe9346d add a few late stragglers.
llvm-svn: 129052
2011-04-07 03:08:22 +00:00
Sean Callanan 6c6be946fc Fixed a bug where missing EDInstInfo would cause
tokenization to crash and burn.

llvm-svn: 129051
2011-04-07 01:56:01 +00:00
Johnny Chen 313ec7953a Sanity check MSRi for invalid mask values and reject it as invalid.
rdar://problem/9246844

llvm-svn: 129050
2011-04-07 01:37:34 +00:00
Eli Friedman c5f22a7815 PR9634: Don't unconditionally tell the AliasSetTracker that the PreheaderLoad
is equivalent to any other relevant value; it isn't true in general.
If it is equivalent, the LoopPromoter will tell the AST the equivalence.
Also, delete the PreheaderLoad if it is unused.

Chris, since you were the last one to make major changes here, can you check
that this is sane?

llvm-svn: 129049
2011-04-07 01:35:06 +00:00
Johnny Chen c0e86fb965 The ARM disassembler was not recognizing USADA8 instruction. Need to add checking for register values
for USAD8 and USADA8.

rdar://problem/9247060

llvm-svn: 129047
2011-04-07 01:05:52 +00:00
Evan Cheng a7c7b54dde Change -arm-divmod-libcall to a target neutral option.
llvm-svn: 129045
2011-04-07 00:58:44 +00:00
Evan Cheng b7c9c407f9 Remove dead code. rdar://9221736.
llvm-svn: 129044
2011-04-07 00:56:37 +00:00
Johnny Chen d4cced54b3 Should also check SMLAD for invalid register values.
rdar://problem/9246650

llvm-svn: 129042
2011-04-07 00:50:25 +00:00
Devang Patel e48ddf863b Simplify. isIdenticalToWhenDefined() checks opcode.
llvm-svn: 129041
2011-04-07 00:30:15 +00:00
Nick Lewycky 561f1755cc Set unnamed_addr on strings created through the IRBuilder.
llvm-svn: 129040
2011-04-07 00:14:29 +00:00
Nick Lewycky 9363fdc9b7 Add support for ArrayRef in IRBuilder's CreateCall.
llvm-svn: 129039
2011-04-07 00:03:25 +00:00
Owen Anderson bdff1c997a Teach the ARM peephole optimizer that RSB, RSC, ADC, and SBC can be used for folded comparisons, just like ADD and SUB.
llvm-svn: 129038
2011-04-06 23:35:59 +00:00
Owen Anderson f9bd6bad8a Cleanups from Jim: remove redundant constraints and a dead FIXME.
llvm-svn: 129036
2011-04-06 22:45:55 +00:00
Devang Patel d715ec82b4 While folding branch to a common destination into a predecessor, copy dbg values also.
llvm-svn: 129035
2011-04-06 22:37:20 +00:00
Jim Grosbach 6ade7e0bac Tidy up.
llvm-svn: 129034
2011-04-06 22:35:47 +00:00
Johnny Chen bd9a4f8d07 A8.6.393
The ARM disassembler should reject invalid (type, align) encodings as invalid instructions.

So, instead of:
Opcode=1641 Name=VST2b32_UPD Format=ARM_FORMAT_NLdSt(30)
 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10  9  8  7  6  5  4  3  2  1  0 
-------------------------------------------------------------------------------------------------
| 1: 1: 1: 1| 0: 1: 0: 0| 0: 0: 0: 0| 0: 0: 1: 1| 0: 0: 0: 0| 1: 0: 0: 1| 1: 0: 1: 1| 0: 0: 1: 1|
-------------------------------------------------------------------------------------------------

	vst2.32	{d0, d2}, [r3, :256], r3

we now have:
Opcode=1641 Name=VST2b32_UPD Format=ARM_FORMAT_NLdSt(30)
 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10  9  8  7  6  5  4  3  2  1  0 
-------------------------------------------------------------------------------------------------
| 1: 1: 1: 1| 0: 1: 0: 0| 0: 0: 0: 0| 0: 0: 1: 1| 0: 0: 0: 0| 1: 0: 0: 1| 1: 0: 1: 1| 0: 0: 1: 1|
-------------------------------------------------------------------------------------------------

mc-input.txt:1:1: warning: invalid instruction encoding
0xb3 0x9 0x3 0xf4
^

llvm-svn: 129033
2011-04-06 22:14:48 +00:00
Jim Grosbach 9c14679260 tidy up.
llvm-svn: 129032
2011-04-06 22:13:52 +00:00
Jakob Stoklund Olesen 8ce2f43694 Also account for the spill code that would be inserted in live-through blocks with interference.
llvm-svn: 129030
2011-04-06 21:32:41 +00:00
Jakob Stoklund Olesen 81439a83f4 Abort the constraint calculation early when all positive bias is lost.
Without any positive bias, there is nothing for the spill placer to to. It will
spill everywhere.

llvm-svn: 129029
2011-04-06 21:32:38 +00:00
Nick Lewycky 469d171ae2 Fix typo in doxy-comment.
llvm-svn: 129028
2011-04-06 20:54:07 +00:00
Johnny Chen 2ac486e387 A8.6.92 MCR (Encoding A1): if coproc == '101x' then SEE "Advanced SIMD and VFP"
Since these "Advanced SIMD and VFP" instructions have more specfic encoding bits
specified, if coproc == 10 or 11, we should reject the insn as invalid.

rdar://problem/9239922
rdar://problem/9239596

llvm-svn: 129027
2011-04-06 20:49:02 +00:00
Nick Lewycky 0cbfcb2b6d Fix comment to use llvm 2.x syntax.
llvm-svn: 129025
2011-04-06 20:38:44 +00:00
Nick Lewycky e30f330393 Replace const std::vector& with ArrayRef in the type creation APIs.
llvm-svn: 129024
2011-04-06 20:28:34 +00:00
Jakob Stoklund Olesen 6895b87dfe Keep track of the number of positively biased nodes when adding constraints.
If there are no positive nodes, the algorithm can be aborted early.

llvm-svn: 129021
2011-04-06 19:14:00 +00:00
Jakob Stoklund Olesen 36b5d8a698 Break the spill placement algorithm into three parts: prepare, addConstraints, and finish.
This will allow us to abort the algorithm early if it is determined to be futile.

llvm-svn: 129020
2011-04-06 19:13:57 +00:00
Roman Divacky 620f6967e0 Fix a typo.
llvm-svn: 129019
2011-04-06 19:12:21 +00:00
Johnny Chen 8bca174f48 Fix a bug in the disassembly of VGETLNs8 where the lane index was wrong.
Also set the encoding bits (for A8.6.303, A8.6.328, A8.6.329) Inst{3-0} = 0b0000,
in class NVLaneOp.

rdar://problem/9240648

llvm-svn: 129015
2011-04-06 18:27:46 +00:00
Rafael Espindola b4dd95b4f9 Add another case we are not optimizing.
llvm-svn: 129012
2011-04-06 17:35:32 +00:00
Rafael Espindola 7a3b244d45 The original issue has been fixed by not doing unnecessary sign extensions.
Change the test to force a sign extension and expose the problem again.

llvm-svn: 129011
2011-04-06 17:19:35 +00:00
Devang Patel 8aa6c326ac face+palm
Keep track of llvm.dbg.value intrinsics with non null values.

llvm-svn: 129010
2011-04-06 17:08:15 +00:00
Rafael Espindola 337a1b29e5 Do a topological sort of the types before writing them out.
This takes the linking of libxul on linux from 6m54.931s to 5m39.840s.

llvm-svn: 129009
2011-04-06 16:49:37 +00:00
Jim Grosbach 300504c135 EngineBuilder setter method for UseMCJIT was missing return value.
llvm-svn: 129008
2011-04-06 16:35:19 +00:00
Chris Lattner 958d299fca add rubinius
llvm-svn: 129006
2011-04-06 16:14:25 +00:00
Frits van Bommel 87e3367d32 Fix a few instances of "warning: extra ';' outside of a function [-pedantic]".
llvm-svn: 129002
2011-04-06 12:29:56 +00:00
Nadav Rotem cc771acd77 This testcase passed even without the fix. Added the target info to make the
test fail (without the fix). Thanks Dan.

llvm-svn: 128999
2011-04-06 11:18:29 +00:00
Duncan Sands ce5d9aedb9 Fix some typos. Minor tweaks to how some things were expressed.
llvm-svn: 128997
2011-04-06 08:07:40 +00:00
Jay Foad 1a7cc44157 Trivial typo fixes.
llvm-svn: 128996
2011-04-06 07:55:30 +00:00
Nick Lewycky d4b3d29c50 Add an empty key for DebugLoc so that you can store an empty DebugLoc in a
DenseMap.

llvm-svn: 128994
2011-04-06 06:49:59 +00:00
Chris Lattner 17c170a8e0 Finish up the first draft of the release notes.
improvements are welcome, please commit any changes directly to SVN.

llvm-svn: 128992
2011-04-06 06:29:50 +00:00
Chris Lattner 23e16b59d2 continue writing.
llvm-svn: 128990
2011-04-06 05:50:04 +00:00
Nick Lewycky ddc72895d0 Support using DebugLoc's in a DenseMap.
llvm-svn: 128988
2011-04-06 05:36:52 +00:00
Jakob Stoklund Olesen f3b2dcc74d Oops. Scary.
llvm-svn: 128986
2011-04-06 04:07:14 +00:00
Jakob Stoklund Olesen bf91c4e85e Analyze blocks with uses separately from live-through blocks without uses.
About 90% of the relevant blocks are live-through without uses, and the only
information required about them is their number. This saves memory and enables
later optimizations that need to look at only the use-blocks.

llvm-svn: 128985
2011-04-06 03:57:00 +00:00
Johnny Chen 0ec0e98a6a Add a missing opcode (SMLSLDX) to BadRegsMulFrm() function.
Add more complete sanity check for LdStFrm instructions where if IBit (Inst{25})
is 1, Inst{4} should be 0.  Otherwise, we should reject the insn as invalid.

rdar://problem/9239347
rdar://problem/9239467

llvm-svn: 128977
2011-04-06 01:18:32 +00:00
Chris Lattner ae6a89a890 add the external users that emailed me.
llvm-svn: 128974
2011-04-06 01:13:49 +00:00
Jim Grosbach 23de2437bd RuntimeDyld should use the memory manager API.
Start teaching the runtime Dyld interface to use the memory manager API
for allocating space. Rather than mapping directly into the MachO object,
we extract the payload for each object and copy it into a dedicated buffer
allocated via the memory manager. For now, just do Segment64, so this works
on x86_64, but not yet on ARM.

llvm-svn: 128973
2011-04-06 01:11:05 +00:00
Chris Lattner 14a3333cb3 some libc++ notes.
llvm-svn: 128970
2011-04-06 00:59:18 +00:00
Chris Lattner 9d5b3712de some edits.
llvm-svn: 128969
2011-04-06 00:56:12 +00:00
Chris Lattner c3a2c9817c distribute content out to the appropriate sections
llvm-svn: 128968
2011-04-06 00:45:11 +00:00
Owen Anderson 867846b1f0 Reapply r128946 (pseudoization of various instructions), and fix the extra imp-def of CPSR it was adding.
llvm-svn: 128965
2011-04-05 23:55:28 +00:00
Chandler Carruth 086f7080d6 Make the virtual destructor out-of-line so we have a key function.
llvm-svn: 128964
2011-04-05 23:54:31 +00:00
Jakob Stoklund Olesen 858afbb6ac Sign error
llvm-svn: 128963
2011-04-05 23:43:16 +00:00
Jakob Stoklund Olesen 5c482cd38f Don't crash when a value is defined after the last split point.
llvm-svn: 128962
2011-04-05 23:43:14 +00:00
Jakob Stoklund Olesen 30b5473d82 Permit blocks to branch directly to a landing pad.
Treat the landing pad as a normal successor when that happens.

llvm-svn: 128961
2011-04-05 23:43:11 +00:00
Jim Grosbach f642c1b600 Remove extraneous 'return'.
llvm-svn: 128959
2011-04-05 23:39:08 +00:00
Johnny Chen f6e327c6a3 Fix a typo in the handling of PKHTB opcode, plus add sanity check for illegal register
encodings for DisassembleArithMiscFrm().

rdar://problem/9238659

llvm-svn: 128958
2011-04-05 23:28:00 +00:00
Chris Lattner ce6b047a10 finished my pass through all the 2.9 commits, now to make this intelligible
llvm-svn: 128955
2011-04-05 23:22:33 +00:00
Chris Lattner a76507fdc3 mention llvm::ArrayRef, which should be use much more pervasively than
it already is.

llvm-svn: 128954
2011-04-05 23:18:20 +00:00
Bob Wilson d135c696c0 Clean up some code for clarity.
llvm-svn: 128953
2011-04-05 23:03:25 +00:00
Chandler Carruth f920fbf94b Provide an empty virtual destructor to go with the virtual methods in
this class. Clang was warning on this with -Wnon-virtual-dtor.

llvm-svn: 128952
2011-04-05 23:03:24 +00:00
Owen Anderson 61e7a935bd Revert r128946 while I figure out why it broke the buildbots.
llvm-svn: 128951
2011-04-05 23:03:06 +00:00
Johnny Chen c3656d29f6 A7.3 register encoding
Qd -> bit[12] == 0
    Qn -> bit[16] == 0
    Qm -> bit[0]  == 0

If one of these bits is 1, the instruction is UNDEFINED.

rdar://problem/9238399
rdar://problem/9238445

llvm-svn: 128949
2011-04-05 22:57:07 +00:00
Devang Patel 9f738849ab Add support to encode function's template parameters.
llvm-svn: 128947
2011-04-05 22:52:06 +00:00
Owen Anderson 3501655ad9 Give RSBS and RSCS the pseudo treatment.
llvm-svn: 128946
2011-04-05 22:42:54 +00:00
Johnny Chen 9da60e016b ARM disassembler was erroneously accepting an invalid RSC instruction.
Added checks for regs which should not be 15.

rdar://problem/9237734

llvm-svn: 128945
2011-04-05 22:18:07 +00:00
Chris Lattner 57ee5a5db7 remove postdom frontiers, because it is dead. Forward dom frontiers are
still used by RegionInfo :(

llvm-svn: 128943
2011-04-05 21:57:17 +00:00
NAKAMURA Takumi 530d739097 docs/ReleaseNotes.html: Please use proper *Heading* elements instead of classified div.
It would be better to browse without stylesheet. (eg. on ViewVC)

llvm-svn: 128942
2011-04-05 21:55:14 +00:00
Johnny Chen 25883487a1 ARM disassembler was erroneously accepting an invalid LSL instruction.
For register-controlled shifts, we should check that the encoding constraint
Inst{7} = 0 and Inst{4} = 1 is satisfied.

rdar://problem/9237693

llvm-svn: 128941
2011-04-05 21:49:44 +00:00
Owen Anderson 77aa266de8 Fix bugs in the pseuo-ization of ADCS/SBCS pointed out by Jim, as well as doing the expansion earlier (using a custom inserter) to allow for the chance of predicating these instructions.
llvm-svn: 128940
2011-04-05 21:48:57 +00:00
Chris Lattner 706754c1f2 remove graphprinter support for domfrontier.
llvm-svn: 128938
2011-04-05 21:43:56 +00:00
Jakob Stoklund Olesen 1ec41e2bd9 These tests no longer require linear scan because reserved register coalescing is now universal.
llvm-svn: 128936
2011-04-05 21:40:41 +00:00
Jakob Stoklund Olesen 6aa0fbf4c0 Run LiveDebugVariables in RegAllocBasic and RegAllocGreedy.
llvm-svn: 128935
2011-04-05 21:40:37 +00:00
Chris Lattner 269ef9f6f0 only 7 more weeks to go.
llvm-svn: 128933
2011-04-05 21:37:08 +00:00
Devang Patel d4e20eacf0 Refactor.
llvm-svn: 128929
2011-04-05 21:08:24 +00:00
Jakob Stoklund Olesen 04123d40cd Document llvm/ADT/IntEqClasses.h in ProgrammersManual.html
llvm-svn: 128927
2011-04-05 20:56:08 +00:00
Bob Wilson 6c20b88173 Add an assertion instead of crashing when the scavenger goes past the end
of a basic block.

llvm-svn: 128925
2011-04-05 20:44:15 +00:00
Nick Lewycky ee54fa29d5 Fix typos. Adjust some whitespace for style. No functionality change.
llvm-svn: 128924
2011-04-05 20:39:27 +00:00
Johnny Chen e9c644d4a0 The r128085 checkin modified the operand ordering for MRC/MRC2 instructions.
Modify DisassembleCoprocessor() of ARMDisassemblerCore.cpp to react to the change.

rdar://problem/9236873

llvm-svn: 128922
2011-04-05 20:32:23 +00:00
Roman Divacky 92d313a021 Add support for detection of Intel SandyBridge.
llvm-svn: 128920
2011-04-05 20:25:36 +00:00
Jakob Stoklund Olesen e20fec7732 Fix one more batch of X86 tests to be register allocation dependent.
llvm-svn: 128919
2011-04-05 20:20:30 +00:00
Jakob Stoklund Olesen 18fd84c79a When dead code elimination removes all but one use, try to fold the single def into the remaining use.
Rematerialization can leave single-use loads behind that we might as well fold whenever possible.

llvm-svn: 128918
2011-04-05 20:20:26 +00:00
Devang Patel 651d06e036 Do not emit empty name.
llvm-svn: 128914
2011-04-05 20:14:13 +00:00
Johnny Chen 151582492d ARM disassembler should flag (rGPRRegClassID, r13|r15) as an error.
llvm-svn: 128913
2011-04-05 19:42:11 +00:00
Andrew Trick 40d388c25e Fix a typo.
llvm-svn: 128912
2011-04-05 19:13:11 +00:00
Andrew Trick 12004013ef Added *hidden* flags -print-options and -print-all-options so
developers can see if their driver changed any cl::Option's. The
current implementation isn't perfect but handles most kinds of
options. This is nice to have when decomposing the stages of
compilation and moving between different drivers. It's also a good
sanity check when comparing results produced by different command line
invocations that are expected to produce the comparable results.

Note: This is not an attempt to prolong the life of cl::Option. On the
contrary, it's a placeholder for a feature that must exist when
cl::Option is replaced by a more appropriate framework. A new
framework needs: a central option registry, dynamic name lookup,
non-global containers of option values (e.g. per-module,
per-function), *and* the ability to print options values and their defaults at
any point during compilation.

llvm-svn: 128910
2011-04-05 18:54:36 +00:00
Johnny Chen 56c15c64b0 LDRD now prints out two dst registers.
llvm-svn: 128909
2011-04-05 18:53:14 +00:00
Johnny Chen 293875ef55 Fix test-llvm failures.
llvm-svn: 128906
2011-04-05 18:41:40 +00:00
Andrew Trick b826ae8310 whitespace
llvm-svn: 128905
2011-04-05 18:41:31 +00:00
Shantonu Sen e59f8ca49e An MCDisassembler has the option to not implement
getEDInfo(), in which case this code would dereference
NULL. EDInst can already handle NULL info, so avoid
the dereference and pass NULL through.

Reviewed by Sean Callanan

llvm-svn: 128904
2011-04-05 18:40:45 +00:00
Jim Grosbach d9dce561b6 Make second source operand of LDRD pre/post explicit.
Finish what r128736 started.

llvm-svn: 128903
2011-04-05 18:40:13 +00:00
Chris Lattner 9ee0b01617 more notes to self.
llvm-svn: 128902
2011-04-05 18:38:45 +00:00
Johnny Chen 33d3a9fadc Constants with multiple encodings (ARM):
An alternative syntax is available for a modified immediate constant that permits the programmer to specify
the encoding directly. In this syntax, #<const> is instead written as #<byte>,#<rot>, where:

    <byte> is the numeric value of abcdefgh, in the range 0-255
    <rot> is twice the numeric value of rotation, an even number in the range 0-30.

llvm-svn: 128897
2011-04-05 18:02:46 +00:00
Johnny Chen 268d63f307 Check for invalid register encodings for UMAAL and friends where:
if dLo == 15 || dHi == 15 || n == 15 || m == 15 then UNPREDICTABLE;
    if dHi == dLo then UNPREDICTABLE;

rdar://problem/9230202

llvm-svn: 128895
2011-04-05 17:43:10 +00:00
Owen Anderson f7678b83d2 Convert ADCS and SBCS instructions into pseudos that are expanded to the ADC/ABC with the appropriate S-bit input value.
llvm-svn: 128892
2011-04-05 17:24:25 +00:00
Stuart Hastings 345094777f ARM doesn't support byval yet. XFAIL this test until it does.
llvm-svn: 128891
2011-04-05 17:16:21 +00:00